2171 Proceedings of the XVI ECSMGE Geotechnical Engineering for Infrastructure and Development ISBN 978-0-7277-6067-8 © The authors and ICE Publishing: All rights reserved, 2015 doi:10.1680/ecsmge.60678 Seismic performance of gravity retaining walls subjected to strong excitation Performance séismique de murs de rétention gravitationnelle soumis à une forte excitation E. Garini 1* , R. Zakkak 1 , I. Anastasopoulos 2 and G. Gazetas 1 1 National Technical University of Athens, Athens, Greece 2 University of Dundee, Scotland, UK * Corresponding Author ABSTRACT Seismic earth pressures on retaining structures are evaluated in practice by the pseudo-static extension of the Coulomb method known as the Mononobe–Okabe approach, in which the inertia force on the soil wedge is included. In reality, earthquake loading is not permanent but varies with time. We examine the dynamic response of a gravity wall in a two-layered soil stratum subjected to horizon- tal strong seismic acceleration. The cohesionless soil layers (retained soil and underlying soil) are modelled inelastically with finite-element discretisation. Their properties are studied parametrically. As excitations we employ a number of idealised wavelets and near-fault ground motions recorded in recent earthquakes. We present results for dynamic earth pressures behind the wall, soil displacements, wall settlement and rotation, and accelerations at several points. The relative importance of sliding versus rocking is explored. Comparisons are made with the Mononobe-Okabe method. RÉSUMÉ Les pressions séismiques de la terre qui agissent sur les murs de soutènement sont évaluées dans la pratique en utilisant l’extension pseudo statique de la méthode connue comme l’approche Mononobe-Okabe, dans le cadre de laquelle les forces d’inertie du sol sont incluses. Pourtant, en réalité, la charge séismique n’est pas permanente, comme c’est supposé dans l’approche pseudo statique, mais elle diffère dans le temps. Cet ouvrage examine la réponse séismique d’un mur de rétention gravitationnelle dans un sol à deux strates, et soumis à de fortes accélérations horizontales de la terre. Nous employons comme excitations quelques ondulations idéalisés et plusieurs ac- célérogrammes enregistrés pendant de récents tremblements de terre. Le sol est composé de deux couches (la couche retenue et la couche inférieure). En plusieurs points, nous présentons des résultats quant de la pression dynamique du sol sur le mur, le déplacement du sol, la rotation du mur, et les accélérations. Une comparaison avec les résultat de la méthode Mononobe-Okabe est également effectuée. 1 INTRODUCTION Gravity retaining walls are the type of retaining walls that rely on their weight to retain the material behind it. The weight of the wall has two roles: resisting the overturning of the wall and causing frictional sliding resistance at the base of the wall on one hand, and experiencing inertia loading on the other. Seismic response of even the simplest retaining wall is a complicated soil-structure interaction problem. The displacements of the wall and the dynamic earth pressures depend on the response of the retained soil, the inertia of the wall itself, the foundation soil and the nature of the input motions. The Mononobe- Okabe method (1926), an extension of Coulomb’s method, is the earliest and most widely used analyti- cal method. It gives the total active thrust acting on the wall by applying a pseudostatic inertial force on the soil wedge. The point of application of the thrust is presumed at 1/3 the height of the wall above its base. This method had been modified and simplified by Seed & Whitman (1970). Richards & Elms (1979) determined permanent (inelastic) outward displace- ments, and Nadim & Whitman (1983) permanent sliding and rotation using the Newmark sliding block concept. Veletsos and Younan (1994) modelled the soil as an elastic medium and obtained elastodynamic solutions. Several other studies have also appeared, among which: AI-Homoud & Whitman (1994), Wu & Prakash (1999), Gazetas et al. (2005), Huang